Various applications call for the sensing and resultant determination of position with respect to rapidly reciprocating members. Exemplary classes of such reciprocating members include pistons in free piston internal combustion engines, hydraulic and pneumatic cylinders and pistons and displacers in free piston Stirling engines (FPSEs) or compressors.
Although traditional internal combustion engines are in widespread use and have many advantages, the exhaust emissions associated with these engines can often exceed acceptable levels. In addition, these engines can be overly noisy and require frequent maintenance. Stirling engines represent improvements in these areas since they are extremely efficient and quiet in operation and they may be configured to generate virtually no emissions whatsoever.
A common application associated with Stirling engines is to generate electrical energy from heat energy by means of a free piston Stirling engine (FPSE) driving a linear alternator. In practical application of these types of arrangements, the power demand of the electrical load varies substantially during normal operation due to variations in load impedance and connection and disconnection of some or all of the loads. In addition, operating temperature and other variables may change during operation so as to result in small and not so small changes in power demand over time.
Attainment of optimal free piston engine alternator plant efficiency over a broad range of electrical power demand and engine temperature conditions requires control of power piston displacement. Not only is it desirable to maintain optimal plant efficiency by continuous adjustment of power piston displacement according to electrical load and engine temperature conditions, but it is also critical to maintain the stroke amplitude of the Stirling engine within safe limits. If the load imposed on an FPSE is suddenly reduced by a substantial amount, the stroke amplitude of the engine will be suddenly correspondingly increased. If the increase in amplitude is great enough, the piston can strike the end of its cylinder resulting in unrecoverable engine failure due to deformation of the piston and possibly the cylinder. On the other hand, if the load is substantially and suddenly increased, the stroke amplitude may be suddenly decreased to the point where the engine can no longer oscillate.
To date, there have been two primary solutions to these potential problems. In the first case, the piston stroke is caused to be essentially self-limiting. According to one approach, the displacer is mechanically driven by the oscillating motion of a linear electric motor. In this way, drive parameters such as displacer stroke amplitude are controlled in order to indirectly control piston stroke amplitude and alternator output voltage. While this approach is generally effective, it does suffer from a number of drawbacks. First, it requires additional electromechanical devices and control circuitry which adds to cost and complexity. Additionally, response time to load variations can be slow due to the inertia of the mechanical parts.
In a second class of solutions to the problem of piston and displacer amplitude control, an active approach may be employed wherein piston and/or displacer position is measured and system control is accomplished based thereon. In the case where an FPSE is driving a linear alternator, the voltage induced in the armature of the alternator may be used to ultimately determine piston position. Typically, this is achieved by obtaining stroke amplitude through armature induced voltage.
This class of solutions also suffers from drawbacks. In particular, there exists a load current dependent voltage drop due to resistance and inductance in the armature windings. Moreover, the strength of the alternator field magnets which directly impacts the induced armature voltage will change with temperature and over time. Resistance and inductance voltage drops must be accounted for in terms of the position calculation since load current is flowing during the actual measurement of the observable alternator terminal voltage. While recoverable magnet strength variation with temperature might also be compensated for if magnet temperature can be estimated, it is difficult to account for variation due to magnet aging. Further, if magnets reach a temperature at which there is a pronounced “knee” in their operating B-H characteristic they may be subject an unrecoverable loss of strength when exposed to a strong armature reaction field. Such weakening is not readily detected and can cause a significant understatement of piston excursion leading to damaging impacts. Unfortunately, in many prior art solutions, the voltage drop and field magnet strength variation is not accounted for and inaccurate data results. Alternatively, in order to account for the voltage drop, additional control circuitry and calculations must be performed.
Many prior art methods for monitoring the position of reciprocating members employ magnetic, electromagnetic and optical sensing technologies. Piston position monitoring solutions employing these prior art methods, which are adequately robust and stable with time and temperature fluctuations may be too costly to implement. In general terms, capacitive sensors are known and it is further known that these sensors may be used to determine lateral positioning including in connection with position sensing for reciprocating members. For example, Moser, in U.S. Pat. No. 4,587,850 discloses an apparatus for detecting and measuring the motion of a piston in a cylinder through the use of a variable impedance comprising a dielectric moving between the fixed electrodes of a capacitive structure. The use of a variable capacitive structure in determining position is based upon the premise that capacitance changes based upon the positional relationship of the electrodes. In most cases, capacitive change is based upon a change in the distance between the two electrodes which is inversely proportionally to the resulting capacitance.
Unfortunately, because the relationship between inter-electrode distance and capacitance is non-linear, difficulties may arise in connection with sensors based upon this relationship. In particular, capacitances may be extremely small in practical application and simultaneous accommodation of such a small variation at one position extreme and a much larger one at the other, due to the inverse non-linear position-capacitance relationship, may result in troublesome inaccuracies in sensor measurement and resulting control operations. As a result, these types of sensors may require costly and complex additional corrective signal processing.
In addition to problems that arise in connection with sensing itself, other difficulties in prior art solutions have arisen with respect to signal processing once sensing has been completed. For instance, small sensor capacitance variations on the order of, for example, plus or minus 25 pF which can result from inter-electrode spacing changes ranging from, for example 1-20 millimeters, can be difficult to transmit and process. This is due to many factors, including, for example, the capacitance error introduced by the fact that the transmitting cables themselves have a baseline capacitance which may vary with cable flexing. Additionally, stray capacitance error is easily introduced and can be very troublesome in connection with signal processing when operational variance is small.